OC English by the numbers, pt. 3

This will be the last batch of statistics on our now-closed English school, this time dealing with attendance and homework.

Part 1, on overall student numbers, joining and quitting rates, and other random statistics

Part 2, on joining and quitting rates by month and by year

And this is part 3.  Guess what this pie chart is!

Screen Shot 2016-06-19 at 6.17.53 PM.png
Blue – present, Green – did homework, Orange – excused absence, Red – unexcused

I took a semi-random sample of 14 adults in 4 classes over the course of a year, and tallied their attendance, homework, and absence rates (unexcused absences are when they didn’t let us know at least 2 hours before the class that they wouldn’t be there).  Just so you know, our classes are at the same time every week – we didn’t offer “flex scheduling” like a lot of the chains do, but if they notified us of an impending absence they were entitled to a makeup in any other class.

As you can see, adults were absent 27% of the time, 13% without notice, with a standard deviation of 24%, meaning there was a wide gap between those who were absent a lot and those who attended most of their scheduled classes – from as high as 79% of classes missed to as low as 2%. We didn’t write down the reasons people were absent even when they gave one, but just based on my recollections, overtime at the office was responsible for the great majority of absences (despite the fact that adult classes were almost always at 8 or 9 pm), followed by forgetting what day of the week it was.  Total absences were negatively correlated with the total number of lessons that student was enrolled for (i.e., before quitting during that period) with a value of -0.34, meaning that many absences somewhat predicted a student’s imminent withdrawal from the class.  Unexcused and excused absences were also correlated with each other with a relatively weak 0.37 (students who missed classes after notifying us were more likely to miss them without notifying us, and vice versa).

TOEIC class statistics were not included in these totals, because that class almost always received either reading or TOEIC-related homework, and was always a small minority of our adult student population.  For that class, attendance was 86% and the homework submission rate was 56%.  This is higher than the average for our adults, but not generalizable to other students or schools because the sample is fewer than 5 people.

Adults turned in homework for about 1/3 of their total scheduled lessons, about half the time they actually attended.  Homework at our school was rewarded with a sticker on a sheet, and when they had gathered 10 stickers they got some sort of prize, usually stationery or snacks.  Doing homework was always an option for adults; we gave unique homework assignments most weeks, and the weeks that we didn’t they always had the option of writing in their diaries or printing out a score sheet from one of the quizzes on their class websites.

I am not sure if other schools give homework, but most schools’ promotional materials and websites do not mention it.  We did because we wanted our students to have some contact with English in the 167 hours of the week that they’re not in class, as well as have something to talk about in the next class.  Plus it gives students an opportunity to see English in something other than a spoken, casual context.

Screen Shot 2016-06-19 at 6.19.36 PM.png

And now we have the breakdown of homework submissions by type in adult classes.  P (present with no homework, blue) is the clear winner here with 53%.  V (video, green), usually with half-sheets for reactions to a Youtube or TED video, was popular as well at 12%.  R (reading, yellow), was almost always articles on the Internet that either I chose and made a special worksheet for or they chose and filled out a generic worksheet, and comprised 7% of submissions.  N (newsletter, orange), was homework usually based on articles I wrote for our monthly newsletter, which gets its own category because I usually gave this homework once a month and is therefore a bit overrepresented at 6% because about one out of four homework assignments were from this category.  W (website, indigo) when I specifically assigned students to take a quiz from their class website, print their score sheet, and bring that in, was turned in only 4% of the time.  O (other, purple) was most often single or half sheets on various topics or language items, a fairly large category which accounted for 17% of all homework submitted.  Homework submission totals for each type were correlated with each other with an average value of 0.62, meaning that students who did one type of homework usually did other types as well.  Total homework submissions for the year ranged between 0 and 32, out of a maximum 47 possible.

What I unfortunately don’t have is statistics on whether students did or didn’t do each type of homework when I assigned it.  The percentages above are simply the rate that homework was turned in the week after I assigned homework of that type.  In rare cases I got a diary notebook turned in after assigning a video or something else like that, but the numbers above assumed homework received is of the type assigned the previous week.  Hence the low value for website homework above, 4%, may reflect fewer website homework assignments, or fewer completions of assigned website homework, or some combination of those two.

I was concerned that individual classes may develop norms of doing or not doing homework, as homework in adult classes was always turned in in full view of everyone, which often led to students complimenting each others’ hardworkingness or laughing when only the newest student did an assignment.  However, when I computed the average rates of completion for each class separately, the results for each class were less divergent than I expected at 15%, 27%, 39%, and 46%; the standard deviation of which is 14%.  Classmates seemed to influence each others’ willingness to do homework to a mild degree.

Next: JHS classes.

Screen Shot 2016-06-19 at 6.20.35 PM.png

Clearly JHS students attended much more often than adults.  Because they still can’t drive themselves to class and they don’t have jobs, they were less likely to skip of their own accord, and their parents were also more likely to let us know when they were going to be absent than to skip classes without giving notice.  A truancy rate this low is a good thing considering that JHS students often have scheduling conflicts with the sports clubs and afterschool classes (juku) they are now likely to be attending.  The correlation between excused absences and unexcused absences was a relatively low 0.19.  Attendance rates ranged between 71% and 100%.

We have no data on homework for JHS students unfortunately, but anecdotally they almost always did their homework, which was much more of a “traditional” reading-and-writing bent.

Last, elementary schoolers.

Screen Shot 2016-06-19 at 9.59.07 PM.png

Even fewer absences.  If there’s a lesson to all this it’s that people are absent from English classes less when their main job is learning, they have no control over their own schedules, and their schedules are fairly regular.  The correlation between the types of absences was higher this time at 0.51, meaning elementary school parents were more likely to be absent without notice if they were often absent with notice. Attendance rates ranged from 83% to 100%.

No homework data for elementary schoolers either, but the same is true of them as for JHS students – they almost always did it or gave it a good try, and it was almost always some sort of reading and writing practice done on paper.

What can other school owners learn from all this?

  • The teacher’s investment of time in designing interesting homework is likely to be paid back in opportunities for interaction with individual students rather than high submission rates.  For a chain school, designing more generic homework assignments would probably be more productive.
  • However, even generic homework should be physically handed out and announced in class.  As I mentioned previously, student use of consistently available resources was very low.
  • Makeup policies for adults need to be designed with higher attrition in mind.  Adults are more likely to miss classes and also to quit, so plan on bringing new students into adult classes fairly frequently and be prepared to let go of habitual truants.  For kids, schools can afford to be more forgiving and focus on maintaining a positive relationship since kids are so infrequently absent.
  • Homework for children can be designed with a little more care since it is much more likely to be completed.
  • Rates of homework completion are more affected by whether that type of homework appeals to individual students, whether it was actually designed individually for students, and how much homework that student usually does than by what class culture may exist.
  • As always, school owners are free to fall on whichever side of the earn/learn dichotomy they wish.  Some schools have clearly decided that all homework (at least the kind you give out for free) is a waste of time, while others devote a lot of outside hours to creating, improving, and correcting it.  What your students expect should probably play a role – as I’ve said before, overshooting your students’ level of dedication with additional services is a recipe for frustration.  If you’re in it for the long haul, try to be the best teacher you can while keeping your efforts efficiently and effectively directed toward whatever purpose you have in mind for your school.  Best of luck to you and your students!

4 thoughts on “OC English by the numbers, pt. 3

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s